矩阵填充结合相似性约束的抗癌药物敏感性预测

Matrix Complecation Combined with Similarity Constraints to Predict Anticancer Drug Sensitivity

为了进行抗癌药物敏感性预测,提出了一种基于矩阵填充与相似性约束的抗癌药物敏感性预测模型——MS模型。该模型综合利用细胞系基因表达数据以及药物化学结构信息,并结合药物与细胞系之间的相似性信息,对抗癌药物敏感性进行预测。将模型应用于CCLE和GDSC这两个基准数据集,十倍交叉验证结果显示:MS模型的预测性能优于部分已发表的经典模型。

In order to predict the sensitivity of anticancer drugs, an new anti-cancer drug sensitivity prediction model, MS model, based on matrix filling and similarity constraint was proposed. Firstly, the model uses the gene expression data of the cell line and the chemical structure information of the drug. Then MS model combines the similarity information between the drug and the cell line. Finally the model achieves the prediction of anticancer drug sensitivity. Applying MS model to two benchmark datasets (CCLE and GDSC), the ten-fold cross-validation results show that the model's prediction performance is better than some of the classic models.